Generative AI on Enterprise Cloud with NiFi and Milvus
What is data science
1. WHAT IS
DATA SCIENCE
AND HOW CAN IT HELP GLOBAL HEALTH?
PART 1 IN A SERIES
DEVELOPED BY JOHN SPENCER
September 2014
h t t p : / / d a t a re v o l u t i o n . u s
2. DATA SCIENCE LETS USERS
IDENTIFY AND UNDERSTAND
PATTERNS IN DATA.
IT BALANCES TRADITIONAL
HYPOTHESIS TESTING AND
PATTERN ANALYSIS.
DATA SCIENCE ALSO
EMPHASIZES MAKING THE
RESULTS OF THE ANALY S I S
EASILY UNDERSTOOD.
F l i c k r i m a g e b y l e e c u l l i v a n
h t t p s : / / f l i c . k r / p / 4 W 5 X y m
3. IT CAN BE A GREAT TOOL FOR
MANAGING AND UNDERSTANDING
COMPLEX DATA .
4. DATA SCIENCE
BRINGS TOGETHER
SKILLS AND METHODS
FROM DIFFERENT
TECHNICAL AND
SUBSTANTIVE AREAS.
MATH AND
STATISTICS
KNOWLEDGE
HACKING
SKILLS
MACHINE
LEARNING
DATA
SCIENCE
DANGER ZONE TRADITIONAL
RESEARCH
SUBSTANTIVE
KNOWLEDGE
FROM DREW CONWAY: b i t . l y / 1 l y G 9 U A
8. NETFLIX
PRIZE
Netflix uses a
recommendation engine to
suggest movies based on
your likes and dislikes.
The machine learning
algorithms that make this
possible rely on data science
principles.
9. MALARIA ATLAS
PROJECT
The researchers at the
Malaria Atlas Project create
models of malaria risk using
Gaussian processes.
Malaria data as well as data
on rainfall, temperature or
land cover are inputs to the
model. The model can help
fill in the gap in areas where
reliable data isn’t available.
www.map.ox.ac.uk
10. ISN’T THAT JUST DATA ANALYSIS?
WHAT MAKES IT DATA SCIENCE?
F l i c k r i m a g e : D e m i - B rooke h t t p s : / / f l i c . k r / p / 4 T n d 2 s
11. TRADITIONAL DATA ANALYSIS
? HY POTHE S I S
QUESTION UNIVERSE OF DATA
ANSWER
!
With traditional data analysis, a hypothesis guides data analysis.
A few data sets are analyzed to prove or disprove the hypothesis.
12. DATA SCIENCE
?
HYPOTHESIS
QUESTION
UNIVERSE OF DATA !
ANSWER
With data science, the data itself can guide analysis. Data
scientists employ a mix of hypothesis testing and pattern
recognition with as many data sets as are relevant.
13. Data science relies on a mix of deductive and inductive
reasoning to create actionable knowledge.
Traditional analysis provides understanding of
phenomenon only where data exists. Data science can
provide understanding where data doesn’t exist.
F l i c k r i m a g e b y f a u n g g h t t p s : / / f l i c . k r / p / 5 n 2 e F r
15. Around the world, there is more data collected associated
with health programs than ever before.
16. Paradoxically, despite the fact there is more data than
before, there are still data gaps.
It is not possible to collect data about every aspect of
health, there will always be data that can’t be collected.
Data science can help mitigate the effect of data gaps.
17. In fact, there is more data about the world in general.
This data can provide valuable information about the
context in which the programs exist.
F l i c k r i m a g e : P o s s i b l e h t t p s : / / f lic.kr/p/eyGbM9
18. In short, there’s more data about the world than ever
before. That includes health related data.
There are still data gaps, things that we don’t know.
Using the data we do have, data science can identify
previously unrecognized patterns and can further our
understanding about things for which data doesn’t exist.
19. Part 1 of a series
Produced by John Spencer
@Jspencerunc
All presentations available via
http://datarevolution.us
Produced under a Creative Commons License